Handler

Warning

Documentation for the module is under development.

Info

This module is intended primarily for developers who want to better understand the structure of the library, as well as those who are going to implement their own modules.

Description

Parent class for all handlers.

Initializing

def __init__(self, mod, onBatchFinish=None, batchsize=128):

Parameters

Parameter Allowed types Description Default
mod Module Trainable neural network -
onBatchFinish callable Function that will be called upon completion of processing of a data batch None
batchsize int Size of a data batch 128

Explanations

-

Methods

handleFromHost

def handleFromHost(self, data, state=None, macroBatchSize=10000, onMacroBatchFinish=None, random=True):

Functionality

-

Parameters

-

Explanations

-

handle

def handle(self, data, state=None, random=True):

Functionality

-

Parameters

-

Explanations

-

getDataSize

@staticmethod
def getDataSize(data):

Functionality

-

Parameters

-

Explanations

-

parseShapeTree

@classmethod
def parseShapeTree(cls, data, onData, auxdata=None):

Functionality

-

Parameters

-

Explanations

-

sliceData

@classmethod
def sliceData(cls, data, idx, batchsize, postSlice):

Functionality

-

Parameters

-

Explanations

-

onMacroBatchStart

def onMacroBatchStart(self, idx, macroBatchSize, state):

Functionality

-

Parameters

-

Explanations

-

onMacroBatchFinish

def onMacroBatchFinish(self, idx, macroBatchSize, state):

Functionality

-

Parameters

-

Explanations

-

handleBatch

def handleBatch(self, batch, idx, state):

Functionality

-

Parameters

-

Explanations

-